I'm starting to think this is a bug in the raid4-5 driver.  I'm seeing this in my kernel log:

Dec 22 21:56:30 faldara kernel: [ 1939.397600] device-mapper: ioctl: device doesn't appear to be in the dev hash table.
Dec 22 21:56:30 faldara kernel: [ 1939.417446] quiet_error: 9 callbacks suppressed
Dec 22 21:56:30 faldara kernel: [ 1939.417455] Buffer I/O error on device dm-9, logical block 0
Dec 22 21:56:30 faldara kernel: [ 1939.417470] Buffer I/O error on device dm-9, logical block 1
Dec 22 21:56:30 faldara kernel: [ 1939.417480] Buffer I/O error on device dm-9, logical block 2
Dec 22 21:56:30 faldara kernel: [ 1939.417489] Buffer I/O error on device dm-9, logical block 3
Dec 22 21:56:30 faldara kernel: [ 1939.417498] Buffer I/O error on device dm-9, logical block 4
Dec 22 21:56:30 faldara kernel: [ 1939.417507] Buffer I/O error on device dm-9, logical block 5
Dec 22 21:56:30 faldara kernel: [ 1939.417516] Buffer I/O error on device dm-9, logical block 6
Dec 22 21:56:30 faldara kernel: [ 1939.417525] Buffer I/O error on device dm-9, logical block 7
Dec 22 21:56:30 faldara kernel: [ 1939.417543] Buffer I/O error on device dm-9, logical block 0
Dec 22 21:56:30 faldara kernel: [ 1939.417552] Buffer I/O error on device dm-9, logical block 1
Dec 22 21:56:30 faldara kernel: [ 1939.419218] device-mapper: table: 252:10: raid45: Invalid RAID device offset parameter
Dec 22 21:56:30 faldara kernel: [ 1939.419230] device-mapper: ioctl: error adding target to table

And then the dmraid process becomes stuck in the uninterruptable state.

Can you check your kernel log for similar entries, and if dmraid -ay becomes stuck, hit alt-sysrq-w and it should add lines to the log that look something like:

Dec 22 21:57:06 faldara kernel: [ 1975.682785] SysRq : Show Blocked State
Dec 22 21:57:06 faldara kernel: [ 1975.682796]   task                        PC stack   pid father
Dec 22 21:57:06 faldara kernel: [ 1975.682864] dmraid        D 0000000100028e87     0  3700   3475 0x00000004
Dec 22 21:57:06 faldara kernel: [ 1975.682874]  ffff880064a77ce8 0000000000000086 ffffffff00000000 0000000000015980
Dec 22 21:57:06 faldara kernel: [ 1975.682883]  ffff880064a77fd8 0000000000015980 ffff880064a77fd8 ffff88006743c4a0
Dec 22 21:57:06 faldara kernel: [ 1975.682892]  0000000000015980 0000000000015980 ffff880064a77fd8 0000000000015980
Dec 22 21:57:06 faldara kernel: [ 1975.682901] Call Trace:
Dec 22 21:57:06 faldara kernel: [ 1975.682918]  [<ffffffff81588175>] schedule_timeout+0x195/0x310
Dec 22 21:57:06 faldara kernel: [ 1975.682931]  [<ffffffff810702f0>] ? process_timeout+0x0/0x10
Dec 22 21:57:06 faldara kernel: [ 1975.682940]  [<ffffffff8158830e>] schedule_timeout_uninterruptible+0x1e/0x20
Dec 22 21:57:06 faldara kernel: [ 1975.682949]  [<ffffffff81071420>] msleep+0x20/0x30
Dec 22 21:57:06 faldara kernel: [ 1975.682958]  [<ffffffff81457dba>] __dm_destroy+0x9a/0x150
Dec 22 21:57:06 faldara kernel: [ 1975.682965]  [<ffffffff81457ea3>] dm_destroy+0x13/0x20
Dec 22 21:57:06 faldara kernel: [ 1975.682974]  [<ffffffff8145d570>] dev_remove+0x90/0x110
Dec 22 21:57:06 faldara kernel: [ 1975.682981]  [<ffffffff8145d4e0>] ? dev_remove+0x0/0x110
Dec 22 21:57:06 faldara kernel: [ 1975.682988]  [<ffffffff8145ddb5>] ctl_ioctl+0x1a5/0x250
Dec 22 21:57:06 faldara kernel: [ 1975.682996]  [<ffffffff8145de73>] dm_ctl_ioctl+0x13/0x20
Dec 22 21:57:06 faldara kernel: [ 1975.683005]  [<ffffffff81162f0d>] vfs_ioctl+0x3d/0xd0
Dec 22 21:57:06 faldara kernel: [ 1975.683011]  [<ffffffff811637e1>] do_vfs_ioctl+0x81/0x340
Dec 22 21:57:06 faldara kernel: [ 1975.683018]  [<ffffffff81163b21>] sys_ioctl+0x81/0xa0
Dec 22 21:57:06 faldara kernel: [ 1975.683026]  [<ffffffff8158a99e>] ? do_device_not_available+0xe/0x10
Dec 22 21:57:06 faldara kernel: [ 1975.683038]  [<ffffffff8100a0f2>] system_call_fastpath+0x16/0x1b

Please confirm if this is the case.